Professional Documents
Culture Documents
Selecting The Proper Mill For Your Product
Selecting The Proper Mill For Your Product
Ultimate strength
Solids Processing
N^
Yield point
\
Selecting the
Proper Mill for
Your Product
Strain
Kerry Johanson
illing is a mechanistically
driven process that hreaks
solid material into smaller
pieces. Selection of the
optimal mill is a daunting task that
often leaves engineers reljng on vendor suggestions to decide which mill
is right for their job. This article addresses a methodology to score and
rank the various mills for use with a
particular material.
In milling, a materials are broken
up due to a few basic breakage mechanisms. Some particles tend to split,
creating smaller particles that are
one-half to one-third the size of the
original particles. This type of breakage is termed fracture. In some cases,
particles break as corners are knocked
off or subsurface cracks form particles
that are considerably smaller than the
original particles. This type of breakage is termed abrasion. Some materials require many impacts or stressstrain events to induce breakage. This
is termed fatigue and can result in
the formation of both small and large
particles. The rate of breakage always
changes during a fatigue-driven event.
Some particles require large strains
before breakage occurs. These large
strains may be due to plastic deformations, such as in the case of plastic,
rubber and ductile metals, separation
of fibers and internal structures as
in biomass, or due to non-linear elastic effects, such as with tissues and
complex polymers.
The premise for this article's mill
ranking comes from the fact that each
mill induces a unique set of impact or
stress-strain events, or both. If a particular material is sensitive to breakage due to stress-strain events, then
Ultimate strength
Elastic limit
Strain
Material properties
For the purpose of this discussion, it
is useful to define some engineering
terms common to the study of material
mechanics. For a material to break, it
must pass through all of the stages on
a stress-strain diagram. When material is either compressed or stretched,
it undergoes an elastic transition.
Relieving stresses under this condition causes a complete rebound in the
strain imposed on the material. At
some point, the stress becomes great
enough that the material cannot sup-
47
Solids Processing
port elastic behavior and must
deform plastically, resulting in
a permanent deformation of
the material. The energy corresponding to the onset of plastic
deformation equals the area
under the stress-strain curve up
to the elastic limit and is called
the resilience {R), as seen in Fig- FIGURE 3. The orientation of a crack FIGURE 4. S-N fatigue curves are used to determine the stress a pure material can withstand
ure 1. Straining the material far is important in determining material
enough will result in the failure breakage behavior
of the material and cause hreakage.
TABLE 2. SCORING FOR
TABLE 1. SCORING FOR
The energy associated with this breakCONE CRUSHER
JAW CRUSHER
age equals the area under the stressCone crusher
Jaw crusher
strain curve up to the rupture strength
Rank for different types of nfiaterials
Rank for different types of materials
and is called toughness (T) (Figure 2).
9
9
Fracture
Fracture
We can define a hrittleness number
3
Fatigue
3
Fatigue
(BR) based on the ratio of the resil2
1
Abrasion
Abrasion
ience to the toughness as descrihed
1
Shear
1
Shear
in Equation (1).
Score on a scale from 1 to 10 (10 best)
BR = (1)
T
A hrittleness numher close to 1 suggests that the material will hreak at
strain conditions very near the elastic
limit. This denotes a hrittle failure,
suggesting that the material may he
suhject to breakage hy impact conditions. If the hrittleness number is
much less than 1, then particles hreak
after significant plastic deformations
and require significant shear to induce hreakage. These particles will
likely he insensitive to hreakage hy
fracture. Particles with a low hrittleness numher may experience fatigue
and can hreak after repeated stressstrain or impact events. Material can
experience repeated stress events and
these tend to open or close cracks in
the particle, depending on the orientation of the crack relative to the direction of impact or application of stress.
Figure 3 illustrates this phenomenon.
If the impact is perpendicular to the
crack, then the crack closes during impact. Conversely, the crack opens if the
impact is in line with the crack.
During every cycle, the region
near the tip of the crack experiences
stress values in excess of the elastic
limit, causing the crack to grow incrementally. There is t5rpically little
to no control over the direction of the
impact or stress application relative
to the direction of the crack, so only
a portion of the impacts or stressloadings actually lead to crack growth
48
Fly-wheel
Northstone
Feeder
49
Feed funnel
MIcronized product outlet
Solids Processing
and abrasion. This type of mill typically handles particles about 400 mm
in diameter and will create particles
around 13 mm in diameter. This mill's
breakage-inducing behavior is ranked
in Table 3, showing that it is a good
choice for products that are sensitive
to fracture.
Impact mill. An impact mill consists
of three zones, as seen in Figure 8. In
the first zone, air enters the mill. This
air is used to classify the milled material after the milling process.
The second zone is the milling zone.
Impact mills have either plates or
pins that rotate at high speed while
particles are fed into the milling stage
through an external feeder. The impact
of the pins or plates causes some of the
material to fracture, creating smaller
particles. If pins are used, then the
angle of the trajectory during impact
can be variable, as some of the impacts
will be glancing blows and some will be
direct-contact blows. Glancing blows
tend to induce abrasion effects while
direct-contact impacts induce fracture
events. There is very little opportunity
for shear in impact mills.
The third zone in the mill is a classifier stage. In this stage, gas is pushed
or drawn out of the mill at a controlled
velocity, entraining particles of the appropriate size with the gas stream. If
particles Eire not small enough to exit
with the carrier gas, they fall back
into the grinding stage and are subject to repeated impacts until they
are small enough to be carried by
the induced currents created by the
gas stream.
These units typically handle particles around 1 to 2 mm in diameter
and can produce particles as small as
5 micrometers (pm). Depending on the
type of impact mill, glancing blows can
comprise 30% of the total impacts, resulting in abrasion events. The other
impacts are nearly all fracture-inducing events. Because of the classification section, this mill can induce many
repeated-impact events. Thus, materials that are sensitive to fatigue can be
used in this type of mill. These observations are summarized in the rankings in Table 4.
Horizontal air-jet m,ill. A horizontal
air-jet mill uses high-velocity air in a
lined, cylindrical chamber to induce
50
Compressed
feed air or
gas Inlet
Vortex finder
Grinding chamber
Compressed
grind air or
gas
Sturtevant
Replaceable
liners
FIGURE 9.
In a horizontal
air-jet mill, a
high-velocity air
stream is used
to breai< apart
materials
Grind air or
gas manifold
ate particles as fines as 0.5 pm. Glancing blows result in abrasion events
and the collision with other particles
results in fracture events. Because of
the classification, this mill can induce
repeated breakage events and materials sensitive to fatigue can be used in
this mill, leading to its score of 8 for
fatigue in Table 5. No shear is possible
in this style of mill, hence its low score
for shear.
70
4f-r _
/yrJ
60
|505 40
1 30
0.10000
1J
"20
tI
10
10
100
-o-
Os
1,000
Particle diameter, \m
^^20 s
10,000
-o-300 s
FIGURE 11. With an impact velocity of 5.4 m/s, the cumulative particle size changes over time
dm, U V
{i,j)-Sj
Broken
O
O O
ooo
ooo "
oooo
oooo
200.0
Intial breakage
400.0
600.0
800.0
Particle size bin, pm
1,000.0 1,200.0
Final breakage
FIGURE 13. The breakage rates (S) are much different after 60
seconds of repeated impacts
mj \-S -m-
Unbroken
0.0
3
4
51
Solids Processing
Breakage for CaO material - initial selectivity
Bin 5
Bin 4
Bin 3
Bin 2
Bin1
4
5
6
7
Attrition particle size bin #
4
5
6
7
Attrition particle size bin #
Bin1
Bin 2
oBin 3
oBin 4
Bin 5
Bin 2
DBin 4
Bin 5
(5)
Time sequence particle-size data for
CaO were used with Equation (5) to
determine the B(x,y) distribution matrix and the S(x) breakage-rate function, shown in Figures 13-15. For the
purposes of this article, B values are
most relevant, since they reveal how
much fracture or abrasion may occur
with the material. However, S values
can give an indication of fatigue. If the
breakage rates (S values) at the beginning of the testing are low and then
increase, then the material is very
sensitive to fatigue events.
The CaO breakage analysis suggests
that lime material is best described
by a dual-rate breakage. Equation (5)
still applies, but must be implemented
in a piecewise linear fashion to combine the solution of two constant-rate
equations. One equation set will consider the first 60 s of breakage and use
the initial conditions at zero seconds.
The other rate equation will consider
breakage after 60 s and use as initial
conditions the particle-size informa-
oBin 3
350
Bin5 = 66 to 123 pm
400
FIGURE 17.
Cumulative particlesize distributions as
a function of time for
lime (CaO) subjected
to repeated impacts
at 5.4 ft/s impact velocities
Bin 6 = 33 to 66 \im
Biii7=16to33Mnl
Initially, almost 70% of the lime breakage is due to fracture with only 10% due
to true abrasion events. However, after
H Fine
I Medium
Fracture
60 s of milling, 33% is due tofractureand
abrasion
abrasion
28% is due to abrasion. Lime is sensitive
FIGURE 16. Abrasion and fracture are
to both fracture and abrasion, but its
the dominant breakage mechanisms for
sensitivity is much greater for fracture
lime (CaO)
events. Now that the modes of breakage
be determined by averaging the respec- have been quantitatively determined,
tive B values. In addition, the average the optimal mill can be selected. In this
fraction that shifts to create the finest case, the optimal mill will be the one
material can also be computed from the where the fracture events induced in
B value data. This gives an indication of the mill are occuring at about 1.5 times
the amoimt of fracture and abrasion the the frequency of the abrasion events. Of
material may be sensitive to (Figure 16). the mills considered in this article, the
Initial
breakage
Final
breakage
Average
breakage
Ingenious Overpressure
Protection for Oii & Gas
KUB
bursting
Unparalleled
cycling capability
98% operating ratio
Torque independent
Leak tight metal
to metal sealing
BliVElUT
I
i
;
'
Optimized sealing
Modular assembly
Maximized corrosion
resistance
Low operational cost
i
'
59929 Brilon/Germany
T.1-49 (0) 29 61 74 05 - 0
F + 49 (0) 29 61 - 5 07 14
sales@rembe.de
53
Solids Processing
As a check of the population-balance
model analysis, the cumulative breakage data points were plotted along with
the computed cumulative breakage
functions from the population-balemce
model (Figure 17). This plot indicates a
good agreement between the breakage
data and the values computed from the
population-balance analysis.
Summary
Mills can be qualitatively ranked based
on their tendency to induce fracture, fatigue, abrasion or shear by examining
the milling action. Mills where classification is part of the milling operation
can induce fatigue events. Mills where
References
1. Epstein, B., Logarithmico Normal Distribution in Breakage of Solids, Ind. Eng. Chem.
Vol 40, pp. 2,281-1,191,1948.
2. Sedlatschek, K. and Bass, L., Contribution
to the Theory of Milling Processes, Powder
Metal. Bull, 6: 148-153., 1953
3. Kapur, P.C. and Agrawal P.K., Approximate
Author
Kerry Johanson is chief
operations officer for Material Flow Solutions, Inc. (7010
NW 23 Way, Suite A, Gainesville, Fla., 32653; Phone; 352379-8879; Cell 352-303-9123;
Email: matflowsol@bellsouth.
net), the consulting firm he
founded in 2001. He also
spends time researching at
the University of Florida Particle Engineering Research
Center. He began his career as a laboratory technician with Jenike & Johanson, before moving
to JR Johanson, Inc. in 1985, where he became
chief technical officer for the company. He has
authored over 40 technical papers, which have
been published in numerous technical journals
internationally and has presented at many industry seminars. He has also developed a graduate course on powder flow and technology at
the University of Florida. Johanson holds a PE
license and has a Ph.D. in chemical engineering
from Brigham Young University.
EASY iNSTALLATION
No holes in tanks or pipes
Away from sensitive processes
VERSATiLE
One size adjusts to motors, from
small up to 150hp
Works on 3 phase, fixed or variable
frequency, DC and single ptiase power
SENSiTIVE
10 times more sensitive than
just sensing amps
CONVENIENT OUTPUTS
For meters, controllers, computers
4-20 miiliamps 0-10 volts
ALL PRODUCTS
MADE IN USA
LOAD CONTROLS
INCORPORATED
'.L0ADC0NTR0LS.COM
Copyright of Chemical Engineering is the property of Access Intelligence LLC d/b/a PBI
Media, LLC and its content may not be copied or emailed to multiple sites or posted to a
listserv without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.